Linear Time Varying Channel Matrix Approach for Modeling MIMO Radar Returns

Conventional approaches to modeling radar returns typically treat clutter stochastically, characterizing clutter returns with a covariance matrix. Such models fail to leverage the waveform dependence of the returns induced by clutter. An alternative model, based on the emerging use of a channel matrix, is set forth and analyzed in this paper. In this formulation, returns from the target and clutter are both modeled as responses of linear time-varying systems. Linear time-invariant systems are intrinsically inadequate to capture the behavior of doubly and triply spread channels. In contrast, the linear time-varying model described here is well suited for modeling clutter and target returns in MIMO radar. We discuss the statistical properties of this model and also present numerical simulations that reveal the effects of spreading.